JP7210030B2 - Methods and kits for diagnosing early pancreatic cancer - Google Patents

Description

INTRODUCTION This patent application claims the benefit of priority to U.S. Provisional Patent Application No. 62/478,860, filed March 30, 2017, the contents of which are hereby incorporated by reference in their entirety. be

Background Pancreatic intraductal papillary mucinous neoplasm (IPMN) is a tumor characterized by intraductal proliferation of neoplastic mucinous cells with varying degrees of cytological atypia, which usually form papillae and pancreatic ducts. lead to cystic dilatation of the cyst and form a clinically detectable mass. Macroscopically, IPMNs are classified into main, complex, and branch duct types based on differential involvement of the pancreatic duct system. Primary and complex IPMNs were more likely to have invasive carcinoma compared with branched tract (48% and 42% vs. 11%), followed by 5-year disease-specificity of primary and complex IPMNs. The survival rate has been shown to be significantly lower than that of the branched tube type (65% and 77% vs. 91%). Histologically, IPMN is thought to progress from low-grade dysplasia (adenoma) to high-grade dysplasia (carcinoma in situ) and invasive carcinoma. Patients with resected non-invasive IPMN have a high 5-year survival rate of 77-94%, while invasive IPMN have a very poor survival rate of 33-43%. Given the significant differences in survival between invasive and noninvasive IPMNs, and between main and branch duct IPMNs, clinical guidelines are critical in determining when lesions should be surgically resected. , has been adopted to assist clinicians. However, while sensitive (97-100%), these guidelines have proven to be highly non-specific (23-30%), especially among branched vessel IPMNs. . Given the prevalence of asymptomatic cysts in an elderly population that tends to have comorbidities, there is a need for more specific tools that can separate high-risk and malignant tumors from low-risk lesions. . In an attempt to improve diagnostic accuracy, analysis of cystic fluid for genetic alterations has been used, including GNAS (guanine nucleotide binding protein (G protein), alpha), KRAS (GTPase KRAS proto-oncogene), IL1B, mucin and microRNAs have been proposed. U.S. Patent Application No. 2017/0022571; Maker, et al. (2011) Ann. Surg. Oncol. 18(1):199; Maker, et al. (2011) Clin. Cancer Res. 17(6):1502- 8; See Maker, et al. (2015) J. Am. Coll. Surg. 220(2):243-253. However, a combination of specific markers of clinically high-risk lesions is needed to aid preoperative diagnosis and risk stratification of patients with IPMN.

SUMMARY OF THE INVENTION The present invention is directed to determining the expression level of one or more biomarker messenger RNA (mRNA) or one or more biomarker microRNA (miRNA) in a sample of pancreatic cyst fluid; by comparing the expression level of one or more biomarker miRNAs with the expression level of one or more biomarkers in a cystic fluid control sample; and classifying the sample as a high risk or low risk intraductal papillary mucinous tumor Methods are provided to distinguish between high-risk and low-risk intraductal papillary mucinous tumors and levels of cystic dysplasia. In one embodiment, the mRNA is ERBB2, GAPDH, GNAS, IL1B, KRAS, MUC-1, MUC-2, MUC-4, MUC-5AC, MUC-7, PGE2-R, PTGER2, PTGES2, PTGS1 and TP63. selected from the group. In another aspect, the miRNA is hsa-miR-101, hsa-miR-106b, hsa-miR-10a, hsa-miR-142-3p, hsa-miR-155, hsa-miR-17-3p, hsa- miR-18a, hsa-miR-21, hsa-miR-217, hsa-miR-24, hsa-miR-30a-3p, hsa-miR-342-3p, hsa-miR-532-3p, hsa-miR- 92a and hsa-miR-99b. In a further aspect, the mRNA or biomarker miRNA are three biomarkers selected from the group hsa-miR-21, hsa-miR-342-3p, IL1B, KRAS, MUC-4, and PTGES2. In particular aspects, the biomarker mRNAs are IL1B, MUC4, and PTGES2. In a still further aspect, the method comprises one or more biomarker proteins such as ERRB2, GAPDH, GNAS, IL1B, KRAS, MUC-1, MUC-2, MUC-4, MUC-5AC, MUC-7, PGE2-R. and measuring the expression of PTGER2. Additionally, the method can include normalizing the relative expression levels of one or more biomarker miRNAs and mRNAs to a reference mRNA, eg, RPLP0. Also provided are kits for identifying levels of high-risk and low-risk intraductal papillary mucinous tumors and cystic dysplasia, comprising one or more biomarkers mRNA or one or more biomarkers in a sample of pancreatic cystic fluid. Contains primer sets for amplifying miRNAs.

Brief description of the drawing
Figure 1 shows a Lasso-penalized logistic regression with cross-validation that identified a three-gene cystic fluid signature with optimal accuracy for predicting pancreatic malignancy risk in IPMN. In this model, low-risk (low and moderate dysplasia) are predicted to high-risk (high-grade dysplasia and invasive carcinoma) cysts with accuracy as measured by an AUC of 86%; y = 0.36 + (-0.06 IL1B) + (-0.17 MUC4) + (-0.50 PTGES2); AUC = 0.86, p-value = 0.002.

DETAILED DESCRIPTION OF THE INVENTION The pancreatic cyst fluid assay is currently being developed as a diagnostic tool to identify patients with premalignant cysts of the pancreas who are at high risk for progression to malignancy. The assay is performed on differentially expressed mRNAs and miRNAs, and optionally proteins, on cystic fluid from patients with pathologically proven low-risk and high-risk (high-grade dysplasia) IPMN. It can be incorporated into a rapid and simple assay that can be analyzed to determine pancreatic cancer risk. The assay of this invention accurately distinguishes high-risk cysts from low-risk cysts with an accuracy of 86%, thereby assisting clinicians in determining when lesions should be surgically resected.

Accordingly, the present invention provides for determining expression levels of one or more biomarker mRNAs and one or more biomarker miRNAs in a sample of pancreatic cyst fluid; expression levels of one or more biomarker mRNAs and one or more biomarker miRNAs compared to the expression level of one or more biomarkers in a serous cystic fluid reference sample; classifying samples; thereby distinguishing between high-risk (invasive and high-grade dysplasia) and low-risk (low-grade and moderate dysplasia) intraductal papillary mucinous tumors.

"Intraductal papillary mucinous neoplasms of the pancreas" or "IPMN" are tumors that grow within the pancreatic ducts (endoducts) and are characterized by the production of a thick fluid by tumor cells (mucinous). Refers to the type of (neoplasm). Intraductal papillary mucinous tumors are important because some of them progress to invasive cancer (benign to malignant transformation) if they are left untreated. Histologic grade of IPMN is based on the highest level of dysplasia present in the lesion. This can be determined by cytological samples obtained from the cyst fluid or wall during EUS-FNA or core biopsy. Criteria for cytological atypia included at least one of the following: increased nucleus-cytoplasmic ratio, increased nuclear size, nuclear compaction, or hyperstaining. Ultimately, this is determined by pathological analysis of permanently prepared surgical pancreatic specimens. Histologic grade is defined as follows: adenoma (dilated pancreatic duct lined with myxoid epithelium, <1 criteria for low-grade dysplasia; also called ductectasia), moderate (>2 criteria for the following: Epithelial tufting, nuclear pseudostratification, nuclear atypia, and mitotic figures; also called borderline), high-grade dysplasia (cribiform or solid growth usually associated with high-grade nuclear atypia; noninvasive also called intraductal carcinoma or carcinoma in situ), and invasive (spread of dysplastic cells into the pancreatic tissue with or without disruption of the ductal basement membrane and lymphatic invasion).

Additional information on the identification, categorization, and characterization of pancreatic lesions, including pancreatic cysts, as currently practiced in surgical techniques, can be found in Current Surgical Therapy, edited by John L. Cameron, 9 ^th ed, 1397 pp, Philadelphia. , PA, Mosby/Elsevier, 2008), and similar texts, reviews, manuals and papers known in the art. Samples for use in the methods of this invention may be obtained by endoscopic ultrasound aspiration, duodenal fluid collection, pancreatic duct aspiration, or direct collection of cystic fluid during surgery.

For the purposes of this invention, "high risk IPMN" refers to invasive and high-grade dysplasia IPMN, while "low risk IPMN" refers to low and moderate dysplasia. Ideally, the method of the invention distinguishes between high-risk IPMN and low-risk IPMN and identifies patients at high risk for progression to malignancy.

A biomarker is an organic biomolecule whose presence in a pancreatic fluid sample is indicated to classify the sample as high-risk IPMN or low-risk IPMN for progression to malignancy. In preferred embodiments, the biomarker is differential in samples taken from subjects with one phenotypic status (e.g., having high-risk IPMN) compared to another phenotypic status (e.g., having low-risk IPMN). expressed When assessed in combination, the biomarkers of this invention are provided for determining whether a subject belongs to one phenotypic status or another. They are therefore useful as markers for disease (diagnosis), therapeutic efficacy of drugs (theranostics), drug toxicity, and selection of appropriate treatment (eg, surgical intervention) for a subject.

mRNA Biomarkers As is customary in the art, mRNA or messenger RNA is a subtype of RNA that encodes the amino acid sequences of proteins. The method of the present invention includes the group ERBB2, GAPDH, GNAS, IL1B, KRAS, MUC-1, MUC-2, MUC-4, MUC-5AC, MUC-7, PGE2-R, PTGER2, PTGES2, PTGS1 and/or TP63. determining or measuring the expression level of 1, 2, 3, 4, 5, 6, 7, 8, 9, or more biomarker mRNAs selected from a sample from the patient. These mRNAs are known in the art under the GENBANK accession numbers shown in Table 1.

In some embodiments, the level of mRNA is at least 20, 30, 40, 50, 60, 80, 90, 100, 110, 120, 130, 140, 150, 160, 170 greater than the reference or control level. , 180, 190, 200, 250, 300, 350, 400, 450, 500, 600, 700, 800, 900, or 1000% higher or lower (or any range derivable therein) if the reference or control level increase or decrease compared to This may or may not involve using standardized or normalized levels of expression in determining whether an increase or decrease is present.

miRNA Biomarkers MicroRNAs (miRNAs) are non-coding RNA molecules approximately 21-23 nucleotides in length that regulate target gene expression by interfering with their transcription or by inhibiting translation. The methods of the present invention include hsa-miR-101, hsa-miR-106b, hsa-miR-10a, hsa-miR-142-3p, hsa-miR-155, hsa-miR-17-3p, hsa-miR- 18a, hsa-miR-21, hsa-miR-217, hsa-miR-24, hsa-miR-30a-3p, hsa-miR-342-3p, hsa-miR-532-3p, hsa-miR-92a and /or the expression level of 1, 2, 3, 4, 5, 6, 7, 8, 9, or more biomarker miRNAs selected from the group of hsa-miR-99b in a sample from the patient Further including determining or measuring. These miRNAs are known in the art under the accession numbers shown in Table 2.

In some embodiments, the level of the miRNA is at least 20, 30, 40, 50, 60, 80, 90, 100, 110, 120, 130, 140, 150, 160, 170 greater than the reference or control level. , 180, 190, 200, 250, 300, 350, 400, 450, 500, 600, 700, 800, 900, or 1000% higher or lower (or any range derivable therein) if the reference or control level increase or decrease compared to This may or may not involve using standardized or normalized levels of expression in determining whether an increase or decrease is present.

Protein Biomarkers In some embodiments, the methods of the invention include ERRB2, GAPDH, GNAS, IL1B, KRAS, MUC-1, MUC-2, MUC-4, MUC-5AC, MUC-7, PGE2-R and/or or determining or measuring in a sample from the patient the expression level of 1, 2, 3, 4, 5, 6, 7, 8, 9, or more biomarker proteins selected from the group of PTGER2 further includes These proteins are known in the art under the GENBANK accession numbers shown in Table 3.

Using the biomarkers described herein, the present invention provides for the expression of one or more biomarker mRNAs and/or one or more biomarker miRNAs and optionally one or more biomarker proteins in a sample of pancreatic cyst fluid. Measuring levels provides a method for identifying, diagnosing and treating high-risk IPMN. Ideally, the expression levels of biomarker mRNAs and miRNAs in samples of pancreatic cyst fluid are measured, thereby providing assays that consist of measuring the expression of nucleic acids (ie, nucleic acid-based). As used in the context of the present invention, a sample of pancreatic cystic fluid comprises cells, nucleic acids, proteins, and/or membrane extracts of cells and is administered to a subject (e.g., human, domestic animal or pet) according to standard clinical practice. animals).

The levels or amounts of biomarker mRNAs and miRNAs in a sample are determined by measuring the levels or amounts of these nucleic acid molecules. Nucleic acid biomarkers include, but are not limited to Northern blot analysis, nuclease protection assay (NPA), linkage analysis of gene expression (SAGE), RNA Seq, in situ hybridization, reverse transcriptase PCR (RT-PCR), PCR , quantitative RT-PCR (qRT-PCR), microarrays, tiling arrays and the like. Because of the ease of use, it is generally desirable to detect nucleic acid molecules using PCR-based approaches. Generally, this involves contacting the sample with two or more PCR primers that specifically hybridize to the biomarker nucleic acid, subjecting the sample to multiple steps of PCR amplification, and determining the amount of amplified sequences ( For example, using gel analysis, blotting methods, incorporation of fluorescently labeled probes and/or fluorescent dyes that intercalate into double-stranded DNA such as SYBR Green). Alternatively, oligonucleotides, aptamers, cDNAs, antibodies or fragments thereof that interact with at least a portion of the biomarker nucleic acids are arrayed on a chip or wafer and used to detect the biomarker nucleic acids. Briefly, these techniques involve methods for rapid and accurate analysis of large numbers of genes. By tagging genes with oligonucleotides or using immobilized probe arrays, chip technology can be employed to separate target molecules as high-density arrays and screen these molecules based on hybridization. can. (See, e.g., Pease, et al. (1994) Proc. Natl. Acad. Sci. USA 91(11):5022-6; Fodor, et al. (1991) Science 251(4995):767-73. ).

Primers (e.g., for PCR-based approaches), probes (e.g., for hybridization-based approaches) or oligonucleotides (e.g., for microarray-based approaches) for use in this aspect are biomarker nucleic acids (see Tables 1 and 2) and generally specifically anneal and amplify at least a portion of the biomarker nucleic acid molecule, but others that encode closely related molecules. do not anneal or amplify the nucleic acid molecules of Suitable primers for amplification of biomarker nucleic acid molecules can be selected by analyzing the sequences provided by the sequences disclosed herein.

Generally, suitable primers are 12-30 bp in length and generate PCR amplification products of 50, 100, 200, 400, 600, 1000 bp or more in length. According to this method, a geometrically amplified product is obtained only if the first and second nucleotide sequences occur within the same biomarker nucleic acid molecule. The basics of non-denaturing PCR are known to those of skill in the art, see, eg, McPherson, et al., PCR, A Practical Approach, IRL Press, Oxford, Eng. (1991).

Exemplary oligonucleotides, forward and reverse primers, or probes for assessing mRNA expression are provided in Table 4. However, other suitable oligonucleotides/primers/probes are well known in the art and available from commercial sources such as Sino Biological, Bio-Rad, OriGene, R&D Systems.

Exemplary oligonucleotides, forward and reverse primers, or probes for assessing miRNA expression are provided in Table 5. However, other suitable oligonucleotides/primers/probes are well known in the art and available from commercial sources such as Sino Biological, Bio-Rad, OriGene, R&D Systems.

Biomarker protein expression is measured in assays using binding agents that specifically bind to the biomarker protein, but not to other proteins. In this embodiment, the sample is contacted with a binding agent (eg, an antibody) that binds to the biomarker protein, and the resulting biomarker-binding agent complexes are detected using standard assays (eg, immunoassays). be. Where the binding agent is, for example, a peptide aptamer, the biomarker-binding agent complex can be directly detected, for example, by a detectable marker protein (such as β-galactosidase, GFP or luciferase) fused to the aptamer. . The level or amount of biomarker-binding agent complex is then correlated with the level of expression of the biomarker protein in the sample.

Binding agents for use in accordance with the invention include antibodies or antibody fragments, and peptide aptamers. In particular aspects of the invention, the binding agents specifically recognize the biomarker proteins listed in Table 3. If the binding agent is an antibody, the antibody can be purchased from commercial sources. Alternatively, antibodies that specifically bind to or recognize a biomarker protein can be generated against antigenic fragments of the biomarker protein. Suitable antigenic regions can be readily identified by those skilled in the art using any of the art-established computer algorithms for identifying such antigenic sequences (e.g., Jamison and Wolf (1988) Bioinformatics 4:181-186). Carmenes, et al. (1989) Biochem Biophys Res Commun. 159(2):687-93).

Various hosts, including goats, rabbits, rats, mice, humans and others, for the production of antibodies are injected with biomarker proteins or any fragments or oligopeptides thereof that have antigenic or immunogenic properties. may be immunized by being Various adjuvants may be used to increase the immune response, depending on the host species. Such adjuvants include, but are not limited to, Freund's, mineral gels such as aluminum hydroxide, and surface active substances such as lysolecithin, pluronic polyols, polyanions, peptides and oil emulsions. Among the adjuvants used in humans, BCG (Bacille Calmette-Guerin) and Corynebacterium parvum are particularly preferred.

Antibodies against biomarker proteins can be generated by immunizing animals with oligopeptides, peptides, or fragments of biomarker proteins. Generally, such oligopeptides, peptides or fragments have an amino acid sequence composed of at least 5 amino acid residues, and more desirably at least 10 amino acid residues. Fragments of biomarker proteins can be generated, for example, by tryptic digestion and extraction from preparative SDS-PAGE gels, or by recombinant fragment expression and purification. In addition, short stretches of amino acids of the biomarker antigen can be fused with those of another protein such as keyhole limpet hemocyanin and antibodies raised against chimeric molecules.

Monoclonal antibodies directed against biomarker proteins can be prepared using any technique that provides for the production of antibody molecules by continuous cell lines in culture. These include, but are not limited to, hybridoma technology, human B-cell hybridoma technology, and EBV-hybridoma technology (Kohler, et al. (1975) Nature 256:495-497; Kozbor, et al. (1985) J. Immunol. Methods 81:31-42; Cote, et al. (1983) Proc. Natl. Acad. Sci. 80:2026-2030; Cole, et al. (1984) Mol. include.

Additionally, antibodies against biomarker proteins include forkhead-associated (FHA) domains, monobodies, minibodies, AFFIBODY molecules, affilins, anticalins, DARPins (i.e., engineered ankyrin repeat proteins), and nanophytins (also called affitins). can be isolated by screening libraries of antibodies or antibody-like molecules such as. Library platforms for screening antibodies or antibody-like molecules include, but are not limited to, phage display (e.g., Benhar & Reiter (2002) Curr. Protoc. Immunol. 48:VI:10.19B:10.19B.1- 10.19B.31), yeast display (see, e.g., Miller, et al. (2005) Prot. Expr. Purif. 42:255-67), and ribosome display (e.g., Douthwaite, et al.). (2006) Prot. Eng. Des. Sel. 19:85-90).

In addition, it uses the splicing of mouse antibody genes into human antibody genes to obtain molecules with appropriate antigen specificity and biological activity, a technique developed for the production of humanized and chimeric antibodies. (Morrison, et al. (1984) Proc. Natl. Acad. Sci. 81, 6851-6855; Neuberger, et al. (1984) Nature 312:604-608; Takeda, et al. (1985) Nature 314:452-454). Alternatively, techniques described for the production of single chain antibodies can be adapted to produce specialized single chain antibodies using methods known in the art. Antibodies with related specificity but distinct idiotypic composition can be generated by chain shuffling from random combinatorial immunoglobulin libraries (Burton (1991) Proc. Natl. Acad. Sci. 88:11120-11123).

Antibodies have also been produced by inducing their production in vivo in lymphocyte populations or by screening immunoglobulin libraries or panels for highly specific binding reagents as is well known in the art. (Orlandi, et al. (1989) Proc. Natl. Acad. Sci. 86: 3833-3837; Winter, et al. (1991) Nature 349:293-299).

Antibodies for use in the methods described herein include, but are not limited to, polyclonal, monoclonal, chimeric, single chain, Fab fragments, bispecific scFv fragments, Fd fragments and Fab expression library generated Contains fragments. For example, fragments include, but are not limited to, F(ab') ₂ fragments that can be produced by pepsin digestion of the antibody molecule and Fab fragments that can be produced by reducing the disulfide bridges of the F(ab') ₂ fragment. . Alternatively, Fab expression libraries can be constructed to allow rapid and easy identification of monoclonal Fab fragments with the desired specificity (Huse, et al. (1989) Science 254:1275-1281).

Diabodies are also contemplated. A diabody isolates the binding domains (both heavy and light chains) of a binding antibody and joins or operably links the heavy and light chains on the same polypeptide chain, thereby performing the binding function. Refers to engineered antibody constructs prepared by supplying linking moieties that conserve (Holliger et al. (1993) Proc. Natl. Acad. Sci. USA 90:6444; Poljak (1994) Structure 2:1121- 1123). This essentially forms a truncated antibody with only the variable domains necessary to bind antigen. A domain is paired with a complementary domain on another chain by using a linker that is too short to allow pairing between the two domains on the same chain, creating two antigen-binding sites. produced. These dimeric antibody fragments or diabodies are bivalent and bispecific. See, for example, Holliger, et al. (1993) supra, Poljak (1994) supra, Zhu, et al. (1996) Biotechnology 14:192-196 and US Pat. No. 6,492,123, incorporated herein by reference. It should be clear that any method of producing diabodies, such as those described in (1999), can be used.

Various immunoassays can be used to measure binding of binding agents to biomarker proteins, and thus to determine expression of biomarker proteins. competitive binding (e.g., ELISA), latex agglutination assays, sandwich immunoassays, gel diffusion reactions, in situ immunoassays, immunoradiometric assays, Western blot analysis, slot blot assays, and any polyclonal or monoclonal antibodies, or fragments thereof Numerous protocols for kinetics (eg, BIACORE™ analysis) using cytoplasm are well known in the art. Such immunoassays typically involve the measurement of complex formation between a specific antibody and its cognate antigen. A two-site, monoclonal-based immunoassay utilizing monoclonal antibodies reactive to two non-interfering epitopes is preferred, but a competitive binding assay can also be employed. See also Methods in cell Biology: Antibodies in Cell Biology (1993) Asai, ed. volume 37; Basic and Clinical Immunology (1991) Stites & Teff, eds. 7th ed.) for a review of immunoassays in general. .

In one embodiment, protein marker analysis involves the use of primary antibodies that specifically bind to the marker protein. In some embodiments, antibody binding is detected by detecting a label on the primary antibody. In another embodiment, the primary antibody is detected by detecting binding of a secondary antibody or reagent to the primary antibody. In further embodiments, the secondary antibody is labeled.

In some embodiments, automated detection assays are used. Methods for automating immunoassays are well known in the art and may be employed in the present invention. In some embodiments, analysis and presentation of results is also automated. For example, in some embodiments, software is used that generates a prognosis based on the presence or absence of a set of proteins corresponding to cancer markers.

Peptide aptamers that specifically bind to biomarker proteins can be rationally designed or screened in libraries of aptamers (eg, provided by Aptanomics SA, Lyon, France). In general, peptide aptamers are synthetic recognition molecules designed based on the structure of antibodies. Peptide aptamers are composed of variable peptide loops attached to both ends of a protein scaffold. This dual structural constraint greatly increases the binding affinity of peptide aptamers to levels comparable to that of antibodies (nanomolar range). Similarly, aptamers that bind to nucleic acid sequences encoding biomarker proteins can also be identified in library screens.

Using the methods of the invention, the expression levels of mRNA and miRNA biomarkers are determined in samples from subjects with pancreatic cancer. To minimize the effects of sample-to-sample variability, methods are usually performed using an internal standard, or one or more reference miRNAs. An ideal internal standard is expressed at consistent levels across different tissues and is unaffected by experimental treatments. RNAs that can be used to normalize patterns of expression include, for example, mRNAs for the reference genes β-actin, GUSB, RPLP0 and TFRC. See, eg, Eisenberg and Levanon (2003) Trends in Genetics 19:362 for a list of additional suitable reference genes. In certain embodiments, the levels of biomarker mRNAs and miRNAs in a sample from a patient are assayed by a quantitative method, and said levels are then "normalized" to the level of mRNA expression of one or more reference mRNAs. ”, which produces normalized expression levels of biomarkers. In some embodiments, the reference mRNA is RPLP0 (Ribosomal Protein Lateral Stalk Subunit P0; GENBANK Accession No. NM_001002).

As an example, mRNA or miRNA levels measured by TaqMan® RT-PCR are referred to as cycle threshold (Ct) values. The lower the Ct, the greater the amount of RNA present in the sample. Expression values of mRNAs or miRNAs in a sample are normalized, eg, by first determining the average expression value at the Ct (Ct _Ref ) of a designated reference mRNA in the sample. A normalized expression value for a biomarker (Ct _biomarker ) is therefore calculated as Ct _biomarker =(C _t )(Ct _Ref ). Optionally, normalized expression values for all biomarkers can be adjusted, eg, such that all adjusted, normalized Cts have values >0. In certain aspects, biomarker expression levels are processed to obtain z-transformed, log2-transformed, normalized (X mean/standard deviation) relative quantification (RQ) values. See, eg, Cheadle, et al. (2003) J. Mol. Diagn. 5:73-81. Additionally, the expression of each of the biomarkers may be weighted based on the biomarker's ability to predict high-risk IPMN or low-risk IPMN.

After the normalized biomarker expression level is determined, it is compared to the expression level of the same biomarkers in serous cyst fluid control samples. Based on these comparisons, if the expression of one or more of the mRNAs and miRNAs in Tables 1 and 2 is increased or decreased (e.g., at least 2-fold or more) compared to the expression level of the same biomarker in the control sample , the sample is classified as high-risk IPMN. Additionally, a sample is classified as low risk IPMN if the expression of one or more of the mRNAs and miRNAs in Tables 1 and 2 is relatively similar compared to the expression levels of the same biomarkers in control samples. In some embodiments, methods of the invention will also include positive and/or negative controls to assess the accuracy of the method.

Alternatively, after normalized biomarker expression levels are obtained, they are compared to predefined expression cutpoints. In certain embodiments, the cutpoints are (a) normalized expression levels and uncertain expression levels that are high-risk IPMNs (i.e., the "upper" cutpoint) and (b) low-risk IPMNs. A numerical boundary is defined between the normalized expression level and the uncertain expression level (ie, the “lower” cutpoint). If the normalized biomarker expression level is not uncertain, the normalized biomarker expression level is clearly designated as either a high-risk IPMN expression level or a low-risk IPMN expression level. Therefore, the sample from which normalized biomarker expression levels were obtained can be designated as a high-risk IPMN sample or a low-risk IPMN sample if it is not in doubt.

As above, the cutpoints are "trained" on preceding samples that are known (via other "validated" methods) to be either high-risk IPMNs or low-risk IPMNs. It is statistically verified in that it is Alternatively, defined cutpoints for an assay can be based on qRT-PCR or other test assays using various statistical tests known in the art. Such statistical tests may include, but are not limited to, Pearson's correlation function, T-test, Mann-Whitney U-test, binomial test, Wilcoxon signed-rank test, analysis of variance and many others.

Once the cutpoints have been determined using the training samples, the test samples are: It is important that they are assayed using the same assay parameters (eg, same primers, amplification conditions, normalization controls, data processing methods, etc.) as the training samples.

The method of this invention helps predict the conversion of premalignant cysts to malignant tumors, thereby providing clinicians with the necessary information to treat patients with IPMN. For example, patients with low-risk lesions may not need surgery and thus avoid the physical, emotional and financial costs of pancreatic resection. By comparison, clinicians can advise patients with high-risk lesions to undergo surgery before pancreatic cancer develops. Accordingly, this invention also provides for prophylactically or therapeutically treating a subject identified as having high-risk IPMN for pancreatic cancer to prevent, delay, ameliorate, slow or reverse the onset or progression of pancreatic cancer. Additional steps of treatment are also provided. Prevention or treatment includes administration of a pharmaceutically effective amount of a therapeutic agent for pancreatic cancer, or one or more other therapies directed to the treatment and/or prevention of pancreatic cancer (e.g., radiotherapy, chemotherapy , immunotherapy, another type of anti-cancer drug, surgery or a combination of two or more of the above). As used herein, the term "preventing" includes avoiding cancer progression and delaying the onset of cancer. In some embodiments, a subject is administered a combination of two or more therapies directed to the treatment of pancreatic cancer. In some embodiments, the subject is treated with a kinase inhibitor.

Additionally, the methods of the invention can include determining biomarker expression at various times after administration of the drug or treatment. A first time (e.g., before drug or treatment delivery) that is higher than a biomarker level detected in a comparable biological sample from the same subject taken at a second time (e.g., after drug or treatment delivery) A biomarker level detected in a biological sample from a subject in is indicative that the subject's pancreatic cancer is regressing.

Kits for measuring the expression of one or more biomarkers disclosed herein are also provided in accordance with the diagnostic and treatment methods of the invention. Kits of the invention include containers containing suitable oligonucleotides or probes for hybridization or nucleic acids encoding one or more biomarker mRNAs of Table 1 and one or more biomarker miRNAs of Table 2. including the primer set of In one embodiment, the kit contains the oligonucleotides or primers provided in Tables 4 and 5. In some embodiments, the kit can further comprise one or more binding agents that bind to the biomarker proteins listed in Table 3. Ideally measure expression of at least 3 of the following markers: IL1B, KRAS, PTGER2, PTGES2, hsa-miR-101, hsa-miR-18a, hsa-miR-217, and hsa-miR-342-3p at least probes or primers for In still other embodiments, the kit can contain a set of primers for amplifying the reference mRNA. In some embodiments, the reference mRNA is RPLP0. Exemplary primers used for amplification of RPLP0 mRNA are forward primer 5'-TGGTCATCCAGCAGGTGTTCGA-3' (SEQ ID NO:58) and reverse primer 5'-ACAGACACTGGCAACATTGCGG-3' (SEQ ID NO:59).

The kit may also contain other solutions and controls necessary or convenient for practicing the invention. The container may be made of glass, plastic or foil and may be a vial, bottle, pouch, tube, bag or the like. Kits also contain written information, such as procedures for practicing the invention, or analytical information, such as the amounts of reagents contained in the first container means, and software for analysis and presentation of results. You may have The container can be in another container, such as a box or bag, along with the written information.

In one embodiment, the kit comprises a solid support such as a chip, microtiter plate or beads, or resin having a capture reagent (e.g., antibody or oligonucleotide) attached thereto, wherein the capture reagent is a biomarker of the invention. , or bind to the nucleic acid that encodes it. The kit also includes instructions for making a wash solution or wash solutions in combination with capture reagents and wash solutions that allow capture of the biomarkers or biomarker nucleic acids on the solid support for subsequent detection. obtain. The kit may contain more than one type of adsorbent each present on a different solid support.

In some embodiments, the present invention provides a barcode-based (e.g., NanoString ( Kits are provided for performing the Trademark)-based) assay. NanoString™-based assays are described in US Pat. No. 8,415,102, US Pat. No. 8,519,115, and US Pat. No. 7,919,237.

Using the kits and methods of the invention, the relative gene expression values may be normalized and transformed before high risk of malignancy (high-grade dysplasia or invasive cancer) or low risk of malignancy (low evaluated in a predictive model to characterize a patient's pancreatic IPMN cysts as having severe or moderate dysplasia). The model may be based on an algorithm in which each gene's expression is weighted based on its ability to predict outcome. Genes that do not contribute to outcome may be excluded to reduce or minimize variability in the algorithm, thereby reducing model complexity and bias.

By way of example, the number of biomarkers in the prediction algorithm may include 6 biomarkers, eg, miR21, miR342, IL1B, KRAS, MUC4, and PTGES2. Using the formula: y = 0.44 + (-0.llmiR217) + (0.03miR342) + (-0.08ILlB) + (0.23KRAS) + (-0.25MUC4) + (-0.85PTGES2) The model has an AUC of 0.82 (p=0.003) and discriminates between high and low risk cysts.

As another example where mutations are present in both Gnas and Kras, the number of biomarkers in the prediction algorithm included eight biomarkers, e.g., miR342, IL1B, KRAS, MUC4, MUC7, PTGER2, PTGES2 and TP63. You can Formula: y = 0.49 + (0.17miR342) + (-0.17IL1B) + (0.18KRAS) + (-0.27MUC4) + (0.02MUC7) + (0.07PTGER2) + (-0.95PTGES2 ) + (0.02TP63), the model has an AUC of 0.82 (p=0.003).

As a further example where mutations are present in either Gnas or KRAS, the number of biomarkers in the prediction algorithm may include three biomarkers, IL1B, MUC4 and PTGES2. Using the formula: y = 0.37 + (-0.06 IL1B) + (-0.01 MUC4) + (-0.50 PTGES2), the model has an AUC of 0.86 (p = 0.002) Predict IPMN to be at high risk.

As yet a further example where there is more than 1 mutation in either Gnas or Kras, the number of biomarkers in the prediction algorithm is 9 biomarkers miR142, miR342, IL1B, KRAS, MUC4, MUC7, PTGER, TP63 and It may contain PTGES2. Formula: y = 0.5l + (0.01miR142) + (0.21miR342) + (-0.l8ILlB) + (0.20KRAS) + (-0.37MUC4) + (0.09MUC7) + (0.16PTGER2) With +(−1.09PTGES2)+(0.07TP63), the model predicts IPMN to be at high risk, with an AUC=0.86 (p=0.002).

In a particular embodiment, the number of biomarkers in the prediction algorithm includes only 3 genes, namely IL1B, MUC4 and PTGES2. Using the formula: y = 0.37 + (-0.06 IL1B) + (-0.01 MUC4) + (-0.50 PTGES2), the model has an AUC of 0.86 (p = 0.002) If y is greater than 0.5, we predict that the IPMN will be at high risk.

Thus, in certain embodiments, the biomarkers used in the methods and kits of the invention are hsa-miR-21, hsa-miR-142-3p, hsa-miR-342-3p, IL1B, KRAS, MUC4, MUC7, PTGER , TP63 and PTGES2. In other embodiments, biomarkers used in the methods and kits of the invention include IL1B, MUC4, and PTGES2. In still other embodiments, the biomarkers used in the methods and kits of this invention consist of IL1B, MUC4, and PTGES2. In further embodiments, the biomarkers used in the methods and kits of this invention include at least IL1B, MUC4, and PTGES2, and may include one or more markers in Tables 1-3.

The invention also provides a sample of pancreatic IPMN cyst fluid and one or more test compounds; and contacting the pancreatic IPMN cyst fluid sample with a test compound; of screening compounds for use in treating pancreatic cancer by detecting changes in the expression of one or more biomarker miRNA, mRNA or protein disclosed herein in a pancreatic cell sample. In some embodiments, the cells are in vitro or in vivo. In other embodiments, candidate compounds are antisense agents (eg, antisense or RNAi molecules) directed against the biomarker miRNAs or mRNAs of the invention. In other embodiments, candidate compounds are antibodies that specifically bind to the biomarker proteins of the invention.

Specifically, the invention provides modulators, i.e., binding to biomarkers of the invention, e.g. having an inhibitory (or stimulatory) effect on biomarker expression or biomarker activity, or e.g. expression of biomarker substrates. or screening methods for identifying candidate or test compounds or agents (eg, proteins, peptides, peptidomimetics, peptoids, small molecules or other drugs) that have a stimulatory or inhibitory effect on the activity of the drug.

Compounds so identified may be useful in the biology of target gene products (e.g., biomarker genes) to modulate the activity of target gene products (e.g., biomarker genes) either directly or indirectly in therapeutic protocols. It can be used to produce functional effects or to identify compounds that interfere with normal target gene interactions. Compounds that inhibit biomarker activity or expression are useful in treating proliferative disorders, such as cancer.

The test compounds of the invention can be used in biological libraries; peptoid libraries (of molecules with novel non-peptide backbones that are resistant to enzymatic degradation yet retain biological activity, but with peptide functionality). spatially addressable parallel solid-phase or solution-phase libraries; synthetic library methods that require deconvolution; "one-bead one-compound" library methods; and affinity It can be obtained using any of a number of approaches in combinatorial library methods known in the art, including synthetic library methods using chromatographic selection. While the biological library and peptoid library approaches are preferred for use with peptide libraries, the other four approaches are applicable to peptide, non-peptide oligomer or small molecule libraries of compounds.

In one embodiment, the assay is a cell-based assay in which cells expressing a biomarker protein or biologically active portion thereof are contacted with a test compound and the ability of the test compound to modulate biomarker activity is determined. be. Determining the ability of a test compound to modulate biomarker activity can be accomplished, for example, by monitoring changes in enzymatic activity. The cells can be of mammalian origin, for example.

In yet another embodiment, the biomarker protein or biologically active portion thereof is contacted with a test compound, and the ability of the test compound to bind to the biomarker protein or biologically active portion thereof is assessed. A cellular assay is provided. Preferred biologically active portions of biomarker proteins for use in assays of the invention include fragments that are involved in interactions with substrates or other proteins, eg, fragments with high surface probability scores.

Cell-free assays involve preparing a reaction mixture of a target gene protein and a test compound under conditions and for a time sufficient to allow the two components to interact and bind, thus forming a complex that can be eliminated and/or detected. including doing Interactions between two molecules can also be detected using, for example, fluorescence energy transfer (FRET). Alternatively, protein ability of biomarkers to bind test compounds can be accomplished using real-time biomolecular interaction analysis (BIA).

Modulators of biomarker expression can also be identified. For example, a cell or cell-free mixture is contacted with a candidate compound and biomarker mRNA or protein expression is assessed relative to the level of biomarker mRNA or protein expression in the absence of the candidate compound. A candidate compound is identified as a stimulator of biomarker mRNA or protein expression if the expression of the biomarker mRNA or protein is greater in the presence of the candidate compound than in its absence. Alternatively, if the expression of the biomarker mRNA or protein is less (i.e., statistically significantly less) in the presence of the candidate compound than in its absence, then the candidate compound is considered an inhibitor of biomarker mRNA or protein expression. identified. The level of biomarker mRNA or protein expression can be determined by methods described herein for detecting biomarker mRNA or protein.

The following non-limiting examples are provided to further illustrate the invention.

Example 1: Single Platform Cyst Fluid Assay for Accurate Prediction of IPMNs with High Malignant Potential for Surgical Resection Biological Samples. The international IPMN cystic fluid collaboration, emerging from the Verona Consensus Conference, consists of a group of high-volume pancreatic surgical centers with IPMN expertise across Europe and the United States (Adsay, et al. (2016) Ann. Surg. 263(1):162-77; Maker, et al. (2015) J. Am. Coll. Surg. 220(2):243-53). IPMN cystic fluid samples were obtained from a future maintained institutional database/repository after institutional review board approval. Only samples with a confirmed diagnosis of IPMN at final pathology and with a specific grade of dysplasia as determined by an expert pancreatic pathologist were included in the study. The most severe dysplasias found in cysts were characterized as low-grade, moderate-grade, high-grade, or invasive carcinomas. As used in multiple other biomarker studies in this field, analyzes were performed for the purpose of risk stratification by specific grades, e.g. Samples were individually assessed by 'risk' (low and moderate dysplasia) or 'high risk' (high-grade dysplasia and invasive carcinoma) pathology (Maker, et al. (2011) Ann. Surg. Oncol. 18(1):199-206; Maker, et al. (2011) Clin. Cancer Res. 17(6):1502-8).

Quantitative analysis of mRNA and miRNA. Total RNA was extracted from 100-400 μl IPMN fluid using Quick-RNA™ MicroPrep R1050/R1051 (Irvine, Calif.) implemented in a Maxwell 16 instrument for automated nucleic acid extraction. DNAse treatment was performed on the instrument according to the manufacturer's instructions. Total RNA was subsequently split into two pathways for mRNA and miRNA analysis using quantitative PCR.

For analysis of mRNA, total RNA was reverse transcribed using random primers and the High-Capacity cDNA Reverse Transcription Kit (Thermo Fisher Scientific) according to the manufacturer's instructions. cDNA was prepared for quantitative PCR (qPCR) using a pre-amplification step using the Taqman® PreAmp Mastermix kit (Thermo Fisher Scientific). Taqman® Gene Expression Assays were pooled to serve as primers for the pre-amplification step according to the manufacturer's instructions.アッセイは、ＩＬ１Ｂ（Ｈｓ０１５５５４１０＿ｍ１）、ＭＵＣ－１（Ｈｓ００１５９３５７＿ｍ１）、ＭＵＣ－２（Ｈｓ００８９４０２５＿ｍ１）、ＭＵＣ－４（Ｈｓ００３６６４１４＿ｍ１）、ＭＵＣ－５ａｃ（Ｈｓ０１３６５６１６＿ｍ１）、ＭＵＣ－７（Ｈｓ００３７９５２９＿ｍ１）、ＰＴＧＥＲ２（Ｈｓ０４１８３５２３＿ｍ１）、ＰＴＧＳ１（ Ｈｓ００３７７７２６＿ｍ１）、ＰＧＥ２－Ｒ（Ｈｓ００１６８７５５＿ｍ１）、ＫＲＡＳ（Ｈｓ００３６４２８２＿ｍ１）、ＧＮＡＳ（Ｈｓ００２５５６０３＿ｍ１）、ＧＡＤＰＨ（Ｈｓ９９９９９９０５＿ｍ１）、ＲＰＬＰ０（Ｈｓ９９９９９９０２＿ｍ１）、ＴＰ６３（Ｈｓ００９７８３４１＿ｍ１）、ＥＲＢＢ２（Ｈｓ０１００１５８０＿ｍ１）、ＰＴＧＥＳ２（Ｈｓ００２２８１５９＿ｍ１）を含んだ。 Pre-amplified cDNA was then used as a template for qPCR reactions with individual assays. Reactions were performed in 384-well plates with Taqman® Fast Advanced Mastermix (Thermo Fisher Scientific) using a ViiA7 real-time PCR instrument (Life Technologies). All reactions were performed in triplicate and in a volume of 10 μl. Real-time data were processed using the comparative C(t) method (Schmittgen & Livak (2008) Nature Protocols 3(6):1101-8). The internal control gene chosen was RPLP0, based on performance across the dataset.

Analysis of miRNAs was performed in a similar manner as mRNAs. Reverse transcription was performed using the Taqman® MicroRNA Reverse Transcription Kit (Thermo Fisher Scientific), using the Taqman® miRNA Assay instead of random primers. The assays used for this study are miR17-3p (hsa-miR-17-3p), miR142-3p (hsa-miR-142-3p), miR532-3p (hsa-miR-532-3p), miR342 -3p (hsa-miR-342-3p), miR30a-3p (hsa-miR-30a-3p), miR21 (hsa-miR-21-5p), miR155 (hsa-miR-155-5p), mir101 (hsa -miR-101-3p), mir10a (hsa-miR-10a-5p), miR106b (hsa-miR-106b-5p), miR18a (hsa-miR-18a-5p), miR217 (hsa-miR-217), It included miR24 (hsa-miR-24-3p), miR92a (hsa-miR-92a-3p), miR99b (hsa-miR-99b-5p), and the gene RNU6B. miRNA qPCR was performed using individual Taqman® miRNA assays as described above for mRNA assays. Real-time data were processed using the comparative C(t) method with the RPLP0 gene as an internal control.

PCR amplification and Sanger sequencing of GNAS and KRAS mutation sites. Genomic DNA was extracted from 100-400 μl of IPMN solution on a Maxwell 16 instrument using the Maxwell 16 Tissue DNA kit (Promega, Madison, Wis.). Mutation analysis of codons 12 and 13 in KRAS and codon 201 in GNAS was performed by PCR followed by Sanger sequencing. Each 50 μl PCR reaction contained 1.5 mM MgCl ₂ , 0.5 μl HotStarTaq® DNA polymerase (Qiagen, Germantown, Md.), 0.2 mM dNTP mix (Sigma-Aldrich Corp., St Louis, Mo.), It contained 1X PCR buffer with 20 pmol of forward and reverse primers and 5 μl DNA template. The primer sequences are as follows: KRAS-F 5'-TGGTGGAGTATTTGATAGTGTATTAACCTTAT-3' (SEQ ID NO: 60), KRAS-R 5'-AAACAAGATTTACCTCTATTGTTGGATCATA-3' (SEQ ID NO: 61), GNAS-F 5'-TCTGAGCCCTCTTTCCAAAACTAC-3' (SEQ ID NO: 61) No. 62), GNAS-R 5'-GGACTGGGGTGAATGTCAAGAA-3' (SEQ ID NO: 63) (Integrated DNA Technologies, Coralville, Iowa). KRAS PCR reactions additionally contained 25 pmol of LNA oligos to suppress wild-type amplification (Exiqon, Woburn, Mass.). After an initial denaturation step at 95°C for 15 min, 40 cycles of PCR were performed as follows: 95°C for 30 sec, 52°C for 30 sec, 68°C, including a final extension step at 68°C for 10 min. for 30 seconds. Amplification products were purified and bi-directionally sequenced on an ABI3130XL gene analyzer using PCR primers and the BigDye 3.1 terminator cycle sequencing kit. Sequence chromatograms were manually visualized to determine if mutations were present. The assay sensitivity is 1% mutant sequence for KRAS codons 12 and 13 and 15% mutant sequence for GNAS codon 201. Appropriate positive and contamination controls were included. Mutation nomenclature followed standard guidelines.

statistical analysis. Cystic fluid gene expression levels were processed to obtain z-transformed, log2-transformed and normalized (X mean/standard deviation) RQ values. The Pearson correlation coefficient was utilized to eliminate highly correlated variables with a cutoff of 0.7. Principal coordinate analysis was then performed. Models were run with additional sequencing data from Kras and Gnas mutation analyzes and evaluated as +kras mutations, +gnas mutations, +gnas/+kras mutations, or 0, 1, or 2 mutations. Mutation analysis as independent variables was added to the data matrix with 22 markers for learning and utilized for classification and regression analysis with support vector machine (SVM) training algorithms. Individual markers were assessed for association with the level of dysplasia. The glmnet package in R was used with logistic regression for sample classification. Batch effect correction was performed. Highly corrected markers were excluded. Lasso-penalized logistic regression with binary classification and 5-fold cross-validation utilized AUC as the criterion for generating optimal signatures.

Selection of targets and cystic fluid. The study includes 14 mRNA markers, 15 miRNA targets, and point mutation analysis of GNAS codon 201 and KRAS codons 12 and 13. A multicenter international IPMN cyst fluid collaboration was conducted and patient samples were contributed to this study. A total of 134 cyst fluid samples were evaluated for inclusion. Adequate fluid volume of samples with specific IPMN pathology and grade of dysplasia (low, moderate, severe, invasive) was identified for 59 cystic fluid samples. 95% of the samples contained sufficient genomic material for further analysis.

Principal component analysis, batch effect correction, elimination of confounders. Principal component analysis demonstrated minimal center bias/clustering with batch effects. Because highly calibrated biomarkers make it difficult to identify individual features for signatures in machine learning algorithms, a Pearson correlation matrix was computed between each pair of markers. Within each group of highly correlated (Pearson correlation>0.7) markers, one representative marker was retained for further analysis using R's caret package. Therefore, confounding markers (ie miR-106B, miR-155, miR-24, miR-532, miR-92A and GNAS) were excluded from the analysis.

Specific Mutation Analysis. GNAS codon 201 and KRAS codons 12 and 13 were sequenced for mutation analysis. Of the 49 samples taken from the cystic fluid with sufficient DNA to reliably sequence, 30 contained point mutations in GNAS or KRAS. For GNAS, 7 samples were p.o. 6 with the R201H mutation and 6 p. have the R201C mutation, while for KRAS, 7 are p. with G12R mutation, 14 p. 8 with the G12V mutation and 8 p. with the G12D mutation, one with the G12F mutation and one with p. with the G12A mutation, one p. had the G13D mutation. Each of the three samples had two KRAS codon 12 point mutations. Nine samples contained both GNAS codon 201 and KRAS codon 12 mutations, one of which also contained a KRAS codon 13 mutation.

は異形成の分類であり、

は、ｉ番目の場合についての（for the i^th case）遺伝子発現（共変量ベクター）である。これは、

を解くためのｌａｓｓｏの目的関数（objective）をもたらし、ここで、ねらいは、高リスクＩＰＭＮと低リスクＩＰＭＮとの間の分類エラーを最小化するマーカーの、最小数であるが最適なサブセットを同定することである（Lockhart, et al. (2014) Ann. Statist. 42(2):413-68；Friedman, et al. (2010) J. Statist. Soft. 33(1):1-22）。 Lasso regression results. After eliminating highly correlated markers, machine learning algorithms were employed to perform feature selection that identified markers significantly associated with the level of dysplasia/risk of pancreatic malignancy. Lasso (Least absolute shrinkage and selection operator) regression analysis modifies the model fitting process to select only a subset of the IPMN grade predictor covariates for use in the final model. performed both variable selection and regularization, which improved the predictive accuracy and interpretability of the regression model. In N patient cystic fluid, each of which consisted of the p-predicted gene and the level of dysplasia as a single outcome;

is a classification of dysplasia,

is the gene expression (covariate vector) for the i ^th case. this is,

yields a lasso objective for solving , where the aim is to identify the minimal but optimal subset of markers that minimizes the classification error between high-risk IPMNs and low-risk IPMNs (Lockhart, et al. (2014) Ann. Statist. 42(2):413-68; Friedman, et al. (2010) J. Statist. Soft. 33(1):1-22).

In a binomial logistic regression model with area under the curve (AUC) as the objective function, the maximum AUC was achieved for miR21, miR342, IL1B, KRAS, MUC4, and PTGES2, to discriminate between low-risk and high-risk cysts. , yielding an AUC of 0.82 (p=0.003). Subset analyzes including repeats, including Gnas and Kras point mutation analysis, were performed to determine the most accurate predictive biosignatures. When considering mutations in either Gnas or Kras, IL1B, MUC4, and PTGES2 achieved the most predictive signatures, y=0.37+(-0.06 IL1B)+(-0.01 MUC4)+( −0.50 PTGES2), AUC=0.86, p-value=0.002 exact formula was constructed (FIG. 1).

Evidence supports a progression model for IPMN from low-grade dysplasia to adenocarcinoma, however, the timeframe for this transformation is unclear. Some lesions progress to cancer, while others may remain low-risk lesions for decades. Ideally, the risk of malignancy is determined preoperatively, but currently many cysts are removed at low risk of malignant transformation at the expense of the patient's mental, physical, and financial costs. with an added risk of approximately 2% mortality and approximately 40% postoperative morbidity. This risk-to-benefit ratio is at the heart of the surgical decision-making challenge in this disease, and the missed spread of potential pancreatic adenocarcinoma, or delayed resection to allow progression to malignancy, is a significant but significant challenge. may result in cancer-related deaths. For this reason, some groups even advocate resection of lesions known to be low risk. In the United States, the majority of patients currently undergoing surgical resection for IPMN are at final pathology despite multiple US, European, and international guidelines for the selection of patients directed to high-risk lesions. would have a low-risk cyst as determined by Up to 65% of lesions predicted by guidelines to be high risk for high-grade dysplasia or invasive cancer were found to be low risk at final pathology, while the same guidelines found low risk for malignant tumors. Other small BD-IPMNs predicted to be at risk have been demonstrated to demonstrate high-risk pathology up to 25% of the time.

The two most commonly used guidelines for clinical decision making in the United States are the revised Sendai (Fukuoka) and the American Gastroenterology Association (AGA) guidelines. Fukuoka has been found to have a high false positive rate with a specificity of 21% for malignancies. The same study found that the AGA guideline has a low false positive rate with a specificity of 44%, but with a high false negative rate and over 12% missed malignancies. A similar analysis of current guidelines supported that Fukuoka had a false-negative rate of 65-72% for identifying high-risk cysts, while AGA misidentified 45% of high-risk IPMNs. Therefore, there is a need in the field for novel and reliable biomarkers that can distinguish cysts with minimal risk of malignant transformation from cysts with high-risk pathology or potential malignancy. there is

In response to this need, an IPMN cyst fluid genetic biosignature has now been developed that has the ability to discriminate high risk with up to 86% accuracy. All high-risk cysts should be surgically resected in otherwise healthy and medically suitable individuals; It can be characterized with this quantitative data that can be used for informed consent.
Bias in this study was mitigated by including an international multicenter cohort, large number of patient cyst fluid samples, running samples in large batches, preselection of candidate biomarkers, and robust statistics including five-fold cross-validation. minimized as much as possible through scientific methods. Interestingly, when kras and gnas mutation analyzes were added to the model, they were not selected as features contributing predictive value, likely due to the high prevalence of these mutations across IPMNs. .

Claims (5)

1. A kit for distinguishing levels of high-risk and low-risk intraductal papillary mucinous tumors and cystic dysplasia, comprising:
comprising one or more primer sets for determining the expression levels of one or more biomarker messenger RNAs (mRNAs) or one or more biomarker microRNAs (miRNAs) in a sample of pancreatic cyst fluid;
Here the primer set is
miR21, miR342, IL1B, KRAS, MUC4, and PTGES2;
miR342, IL1B, KRAS, MUC4, MUC7, PTGER2, PTGES2, and TP63;
IL1B, MUC4, and PTGES2; and/or
miR142, miR342, IL1B, KRAS, MUC4, MUC7, PTGER, TP63 , and PTGES2,
wherein the primer set compares the expression level of one or more biomarkers in the sample to the expression level of one or more biomarkers in a cystic fluid control sample, and classifies the sample as a high-risk or low-risk intraductal papillary mucinous tumor. Said kit, also used to determine the expression level of one or more biomarkers in a cyst fluid control sample for classification. 2. The kit of claim 1, further comprising one or more binding agents for measuring expression of one or more biomarker proteins. one or more biomarker proteins are selected from the group ERRB2, GAPDH, GNAS, IL1B, KRAS, MUC-1, MUC-2, MUC-4, MUC-5AC, MUC-7, PGE2-R, and PTGER2 , the kit of claim 2 . 2. The kit of claim 1, further comprising a primer set for amplifying reference mRNA. 5. The kit of claim 4 , wherein the reference mRNA is RPLP0.

Images